Update app.py

app.py

@@ -1,29 +1,33 @@
-import os
 import gradio as gr
+from fastai.vision.all import *
 import copy
+import os
 import time
-import llama_cpp
 from llama_cpp import Llama
-from huggingface_hub import hf_hub_download  
-
+from huggingface_hub import hf_hub_download
 
+# Load the LLM model
 llm = Llama(
     model_path=hf_hub_download(
         repo_id=os.environ.get("REPO_ID", "TheBloke/Llama-2-7B-Chat-GGML"),
         filename=os.environ.get("MODEL_FILE", "llama-2-7b-chat.ggmlv3.q5_0.bin"),
     ),
     n_ctx=2048,
-    n_gpu_layers=50, # change n_gpu_layers if you have more or less VRAM 
-) 
+    n_gpu_layers=50,  # change n_gpu_layers if you have more or less VRAM
+)
 
 history = []
 
 system_message = """
-You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe.  Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.
-If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
-"""
+You are a helpful BIRD ASSISTANT, AND YOU KNOW A LOT ABOUT BIRDS"""
+# The rest of the system message
 
+# Load the Vision Model
+learn = load_learner('export.pkl')
 
+labels = learn.dls.vocab
+
+# Function to generate responses using LLM
 def generate_text(message, history):
     temp = ""
     input_prompt = f"[INST] <<SYS>>\n{system_message}\n<</SYS>>\n\n "
@@ -36,13 +40,13 @@ def generate_text(message, history):
         input_prompt,
         temperature=0.15,
         top_p=0.1,
-        top_k=40, 
+        top_k=40,
         repeat_penalty=1.1,
         max_tokens=1024,
         stop=[
-            "<|prompter|>",
-            "<|endoftext|>",
-            "<|endoftext|> \n",
+            "",
+            "",
+            " \n",
             "ASSISTANT:",
             "USER:",
             "SYSTEM:",
@@ -56,16 +60,38 @@ def generate_text(message, history):
 
     history = ["init", input_prompt]
 
+# Function to predict using the Vision Model and interact with LLM
+def predict(img):
+    img = PILImage.create(img)
+    pred, pred_idx, probs = learn.predict(img)
+    bird_predictions = {labels[i]: float(probs[i]) for i in range(len(labels))}
+    
+    # Construct a message for LLM using bird predictions
+    message = "I have detected:\n"
+    for bird, prob in bird_predictions.items():
+        message += f"- {bird}: {prob:.2%}\n"
+    
+    # Generate responses using LLM
+    responses = generate_text(message, history)
+    
+    return {"bird_predictions": bird_predictions, "llm_responses": responses}
 
-demo = gr.ChatInterface(
-    generate_text,
-    title="llama-cpp-python on GPU",
-    description="Running LLM with https://github.com/abetlen/llama-cpp-python",
-    examples=["tell me everything about llamas"],
-    cache_examples=True,
-    retry_btn=None,
-    undo_btn="Delete Previous",
-    clear_btn="Clear",
-)
-demo.queue(concurrency_count=1, max_size=5)
-demo.launch()
+title = "Bird Detector with LLM"
+description = "Detect birds and interact with LLM."
+examples = ['BIRD.jpg']
+interpretation = 'default'
+enable_queue = True
+
+gr.Interface(
+    fn=predict,
+    inputs=gr.inputs.Image(shape=(512, 512)),
+    outputs={
+        "bird_predictions": gr.outputs.Data(type="dictionary"),
+        "llm_responses": gr.outputs.Textbox(),
+    },
+    title=title,
+    description=description,
+    examples=examples,
+    interpretation=interpretation,
+    enable_queue=enable_queue,
+).launch()